BCR repertoire sequencing 2x300 Miseq (phiX% at 30) but low Q30 scores

[fjparan]

We recently sequenced BCR repertoire samples prepared using NEBNext Immune Seq Kit, 22 libraries multiplexed in one lane. However, the sequencing results returned poor data with most libraries averaging around 20% Q30 scores. I tried to play around with it by assembling and filtering the pairs using pRESTO and seems like around 1-150 bp could pass Q20, but 151-300 bp is bad. The R2 fastq files are also much poorer in quality than the R1 files

Anyway, I am trying to isolate the problem as this could be a library problem? the PhiX% too high at 30% (Azenta sequencing suggested 30% for immune repertoire)? Or a MiSeq problem (as I have read in this forum complaining about MiSeq)?

I have attached files for your reference. Hoping you could help me understand this dilemma and what steps we can pursue after this.

1. these are some of my bioanalyzer results. granted, one of the libraries H46 may have higher fragments (still included in the library with no additional purification)

2. Azenta QC results with fragments above 700 bp (gel and tapestation)

3. FastQC results of R1 and R2

4. FastQC of assembled pairs

Looking forward to your comments. Let me know where I did wrong, if I can still recover anything from these sequences, and future steps.

PS. this is my first time performing bulk RNA sequencing, but had experience with scRNA and RADseq library preparations.